基于SPSS统计软件的医用装备诊疗检查需求预测的ARIMA模型构建与应用研究  

Construction and application of ARIMA model for medical equipment diagnosis and treatment demand prediction based on SPSS statistical software

作　　者：刘伟军[1] 张千彧 李帅帅[2] 王凤[2] 张恩科[1] LIU Wei-jun;ZHANG Qian-yu;LI Shuai-shuai(Department of Medical Equipment,Shaanxi Provincial People’s Hospital,Xian 710068,China;不详)

机构地区：[1]陕西省人民医院医学装备部,陕西西安710068 [2]陕西省人民医院耗材试剂部,陕西西安710068 [3]西安交通大学生命科学与技术学院生物医学工程系,陕西西安710049

出　　处：《中国医学装备》2020年第9期135-139,共5页China Medical Equipment

基　　金：陕西省创新能力支撑计划(2017KCT-36)“医院临床医学工程创新团队”;陕西省创新能力支撑计划(2019KRM115)“基于方法论和数学模型的大型医用设备的配置优化研究”;陕西省软科学研究计划(2016KRM136)“国产医疗设备在基层医院推广使用策略研究”;西安市科技计划[2019114613YX001SF043(7)]“基于门诊医技资源合理调度的预约排程优化的方法体系研究”。

摘　　要：目的:构建医用装备诊疗检查需求预测的差分自回归移动平均(ARIMA)模型,为管理决策提供参考依据。方法:以2013-2018年医院核磁共振(MR)诊疗检查数据为基础序列,应用SPSS统计软件建立ARIMA模型,运用拟合评价指标对预测模型进行拟合评价。选取2019年MR检查数据对预测数据进行精度检验,预测2020年MR检查数据。结果:构建的预测模型描述为ARIMA(0,1,1)(0,1,0)12,该模型拟合评价总体良好;2019年的数据精度检验评价合理且检验合格,预测数据在置信区间为95%的相对误差控制在-5.74%~5.82%,预测精度良好。该模型可用于2020年MR检查需求预测。结论:SPSS统计软件的运用简化了预测模型的构建、评价和应用,构建的ARIMA预测模型和预测数据可作为医疗机构进行医技排程统筹管理、医技资源整体调度、装备配置合理优化和诊疗服务满意提升的决策参考。Objective: To construct an autoregressive integrated moving average(ARIMA) model which could predict the demand of the examination of diagnosis and treatment of medical equipment so as to provide reference for managerial decision. Methods: The data of diagnosis and treatment of magnetic resonance(MR) of hospital during 2013 year and 2018 year were used as basic sequence, and SPSS software was used to establish ARIMA prediction model, and the fitting evaluation indicators were used to implement fitting evaluation for prediction model. The MR examination data in 2019 year were selected to implement accuracy test for prediction data so as to predict MR examination data of 2020 year. Results: The constructed prediction model was described as ARIMA(0,1,1)(0,1,0)12. The fitting evaluation of this model was generally favorable. The accuracy test and evaluation of these data in 2019 were reasonable and qualified. The relative error of prediction data in the confidence interval with 95% was controlled within-5.74% and 5.82%, and the prediction accuracy of this model was favorable. And the model could be used in the prediction of the demand of MR examination in 2020 year. Conclusion: The application of SPSS software simplifies the construction, evaluation and application of the prediction model. The constructed ARIMA prediction model and prediction data can be used as a decision reference for medical institutions to carry out overall management of the scheduling of medical technique, overall dispatch of the resource of medical technique, reasonable optimization of equipment configuration, and the improvement of the satisfaction of medical service.

关 键 词：SPSS统计软件 差分自回归移动平均(ARIMA)模型 医用装备 需求预测 模型构建 

分 类 号：R197.39[医药卫生—卫生事业管理]